An application of deep learning in the analysis of stellar spectra

Abstract

Spectroscopic surveys require fast and efficient analysis methods to maximize their scientific impact. Here we apply a deep neural network architecture to analyze both SDSS-III APOGEE DR13 and synthetic stellar spectra. When our convolutional neural network model (StarNet) is trained on APOGEE spectra, we show that the stellar parameters (temperature, gravity, and metallicity) are determined with similar precision and accuracy as the APOGEE pipeline. StarNet can also predict stellar parameters when trained on synthetic data, with excellent precision and accuracy for both APOGEE data and synthetic data, over a wide range of signal-to-noise ratios. In addition, the statistical uncertainties in the stellar parameter determinations are comparable to the differences between the APOGEE pipeline results and those determined independently from optical spectra. We compare StarNet to other data-driven methods; for example, StarNet and the Cannon 2 show similar behaviour when trained with the same datasets, however StarNet performs poorly on small training sets like those used by the original Cannon. The influence of the spectral features on the stellar parameters is examined via partial derivatives of the StarNet model results with respect to the input spectra. While StarNet was developed using the APOGEE observed spectra and corresponding ASSET synthetic data, we suggest that this technique is applicable to other wavelength ranges and other spectral surveys.

SPCANet: Stellar Parameters and Chemical Abundances Network for LAMOST-II Medium Resolution Survey

“…Artificial neural network (ANN) methods were firstly adopted to determine stellar atmospheric parameters by Bailer- Jones et al (1997), and rejuvenated recently because of development of new training techniques and hardware. Inspired by the successful application of con- volutional neural networks (CNN) to APOGEE spectra (Fabbro et al 2018;Leung & Bovy 2019), we design a specific CNN structure for transferring stellar labels from APOGEE-payne catalog to LAMOST-II MRS spectra.…”

“…Researchers made efforts on machine learning in stellar parameter estimation, such as The Cannon (Ness et al 2015;Casey et al 2016), The Payne , StarNet (Fabbro et al 2018), AstroNN (Leung & Bovy 2019) and GSN (Wang et al 2019a), and most of them employ artificial neural networks for building regression map relationship. These methods depend on training and test sets usually called reference sets, and the more complete the parameter space covers, the more information can be obtained by the model training.…”

Application of convolutional neural networks for stellar spectral classification

“…m and c denote the slope and intercept of the linear fit. Another parameter that we use to evaluate the goodness of prediction is R 2 -Score (Steel & Torrie 1960;Glantz 1990;Draper 1998), also referred to as coefficient of determination, which indicates the proportion of variance in the predicted labels governed by the expected labels. R 2 -Score can take the values in the range 0-1, and is defined as:…”

“…In this work, we apply shallow neural networks (NNs) as well as deep convolutional neural networks to study optical stellar spectra in detail and investigate whether using deeper networks of convolutional layers can significantly reduce the error and accuracy achieved in the stellar spectral classification. Deep learning frameworks (Hinton & Salakhutdinov 2006;Bengio 2009;Zeiler & Fergus 2013) have been used in the astronomical domain for various applications like galaxy morphology prediction (Dieleman et al 2015), classification of variable stars based on their light curves (Mahabal et al 2017), estimating atmospheric parameters using stellar spectra (Fabbro et al 2018), detecting bar structures in galaxy images (Abraham et al 2018), classifying galaxy morphologies at radio wavelengths (Wu et al 2019) etc. However, all these problems require a large sample for supervised training of the network.…”

Learning the relationship between galaxies spectra and their star formation histories using convolutional neural networks and cosmological simulations

“…Convolutional Neural Networks (CNNs) 3 are growing in popularity in many areas of Astronomy, typically as a means of analysing 2D image data (e.g. Tuccillo et al 2017;Petrillo et al 2017), and have been shown to perform remarkably well, with prediction accuracies in classification tasks approaching human level (Flamary 2016;Fabbro et al 2018).…”